PM 566 Lab 11

Author

Dana Gonzalez

Part 1

Step 1

# Load Libraries
library(dplyr)
library(tidyverse)
library(ggplot2)
library(plotly)
library(zoo)
library(tidyr)

# Read-in and Merge NYT Data
cov_states <- as.data.frame(read.csv("https://raw.githubusercontent.com/nytimes/covid-19-data/master/us-states.csv"))

state_pops <- as.data.frame(read.csv("https://raw.githubusercontent.com/COVID19Tracking/associated-data/master/us_census_data/us_census_2018_population_estimates_states.csv"))
state_pops$abb <- state_pops$state
state_pops$state <- state_pops$state_name
state_pops$state_name <- NULL

cov_states <- merge(cov_states, state_pops, by="state")

Step 2

# Inspect Data
dim(cov_states)
[1] 58094     9
head(cov_states)
    state       date fips   cases deaths geo_id population pop_density abb
1 Alabama 2023-01-04    1 1587224  21263      1    4887871    96.50939  AL
2 Alabama 2020-04-25    1    6213    213      1    4887871    96.50939  AL
3 Alabama 2023-02-26    1 1638348  21400      1    4887871    96.50939  AL
4 Alabama 2022-12-03    1 1549285  21129      1    4887871    96.50939  AL
5 Alabama 2020-05-06    1    8691    343      1    4887871    96.50939  AL
6 Alabama 2021-04-21    1  524367  10807      1    4887871    96.50939  AL
tail(cov_states)
        state       date fips  cases deaths geo_id population pop_density abb
58089 Wyoming 2022-09-11   56 175290   1884     56     577737    5.950611  WY
58090 Wyoming 2022-08-21   56 173487   1871     56     577737    5.950611  WY
58091 Wyoming 2021-01-26   56  51152    596     56     577737    5.950611  WY
58092 Wyoming 2021-02-21   56  53795    662     56     577737    5.950611  WY
58093 Wyoming 2021-08-22   56  70671    809     56     577737    5.950611  WY
58094 Wyoming 2021-03-20   56  55581    693     56     577737    5.950611  WY
str(cov_states)
'data.frame':   58094 obs. of  9 variables:
 $ state      : chr  "Alabama" "Alabama" "Alabama" "Alabama" ...
 $ date       : chr  "2023-01-04" "2020-04-25" "2023-02-26" "2022-12-03" ...
 $ fips       : int  1 1 1 1 1 1 1 1 1 1 ...
 $ cases      : int  1587224 6213 1638348 1549285 8691 524367 1321892 1088370 1153149 814025 ...
 $ deaths     : int  21263 213 21400 21129 343 10807 19676 16756 16826 15179 ...
 $ geo_id     : int  1 1 1 1 1 1 1 1 1 1 ...
 $ population : int  4887871 4887871 4887871 4887871 4887871 4887871 4887871 4887871 4887871 4887871 ...
 $ pop_density: num  96.5 96.5 96.5 96.5 96.5 ...
 $ abb        : chr  "AL" "AL" "AL" "AL" ...

Step 3

# Format Data

# Format Date Variable
cov_states$date <- as.Date(cov_states$date, format="%Y-%m-%d")

#Format State and Abbreviation Variables
state_list <- unique(cov_states$state)
cov_states$state <- factor(cov_states$state, levels = state_list)
abb_list <- unique(cov_states$abb)
cov_states$abb <- factor(cov_states$abb, levels = abb_list)

# Sort by State and Date Variables
cov_states = cov_states[order(cov_states$state, cov_states$date),]

# Re-Inspect Data
str(cov_states)
'data.frame':   58094 obs. of  9 variables:
 $ state      : Factor w/ 52 levels "Alabama","Alaska",..: 1 1 1 1 1 1 1 1 1 1 ...
 $ date       : Date, format: "2020-03-13" "2020-03-14" ...
 $ fips       : int  1 1 1 1 1 1 1 1 1 1 ...
 $ cases      : int  6 12 23 29 39 51 78 106 131 157 ...
 $ deaths     : int  0 0 0 0 0 0 0 0 0 0 ...
 $ geo_id     : int  1 1 1 1 1 1 1 1 1 1 ...
 $ population : int  4887871 4887871 4887871 4887871 4887871 4887871 4887871 4887871 4887871 4887871 ...
 $ pop_density: num  96.5 96.5 96.5 96.5 96.5 ...
 $ abb        : Factor w/ 52 levels "AL","AK","AZ",..: 1 1 1 1 1 1 1 1 1 1 ...
head(cov_states)
       state       date fips cases deaths geo_id population pop_density abb
1029 Alabama 2020-03-13    1     6      0      1    4887871    96.50939  AL
597  Alabama 2020-03-14    1    12      0      1    4887871    96.50939  AL
282  Alabama 2020-03-15    1    23      0      1    4887871    96.50939  AL
12   Alabama 2020-03-16    1    29      0      1    4887871    96.50939  AL
266  Alabama 2020-03-17    1    39      0      1    4887871    96.50939  AL
78   Alabama 2020-03-18    1    51      0      1    4887871    96.50939  AL
tail(cov_states)
        state       date fips  cases deaths geo_id population pop_density abb
57902 Wyoming 2023-03-18   56 185640   2009     56     577737    5.950611  WY
57916 Wyoming 2023-03-19   56 185640   2009     56     577737    5.950611  WY
57647 Wyoming 2023-03-20   56 185640   2009     56     577737    5.950611  WY
57867 Wyoming 2023-03-21   56 185800   2014     56     577737    5.950611  WY
58057 Wyoming 2023-03-22   56 185800   2014     56     577737    5.950611  WY
57812 Wyoming 2023-03-23   56 185800   2014     56     577737    5.950611  WY
# Inspect Ranges
min(cov_states$date)
[1] "2020-01-21"
max(cov_states$date)
[1] "2023-03-23"
min(cov_states$cases)
[1] 1
max(cov_states$cases)
[1] 12169158

The range of dates for this data set is between January 21st, 2020 (01/21/2020) and March 23rd, 2023 (03/23/2023). The minimum number of cases observed is 1, whereas the maximum is 12,169,158.

Step 4

# Add "new_cases" and "new_deaths" and Correct Outliers

### Add New Variables
for (i in 1:length(state_list)) {
  cov_subset <- subset(cov_states, state == state_list[i])
  cov_subset <- cov_subset[order(cov_subset$date), ]
  
  cov_subset$new_cases <- c(0, diff(cov_subset$cases))
  cov_subset$new_deaths <- c(0, diff(cov_subset$deaths))
  
  for (j in 2:nrow(cov_subset)) {
    cov_subset$new_cases[j] = cov_subset$cases[j] - cov_subset$cases[j - 1]
    cov_subset$new_deaths[j] = cov_subset$deaths[j] - cov_subset$deaths[j - 1]}

  cov_states$new_cases[cov_states$state == state_list[i]] <- cov_subset$new_cases
  cov_states$new_deaths[cov_states$state == state_list[i]] <- cov_subset$new_deaths}

### Focus on Recent Dates
cov_states <- cov_states |> dplyr::filter(date >= "2021-06-01")

### Inspect Outliers
p1<-ggplot(cov_states, aes(x = date, 
                          y = new_cases, 
                          color = state)) + 
  geom_line() + 
  geom_point(size = .5, alpha = 0.5)

ggplotly(p1)
p1<-NULL

p2<-ggplot(cov_states, aes(x = date, 
                          y = new_deaths, 
                          color = state)) + 
  geom_line() + 
  geom_point(size = .5, alpha = 0.5)

ggplotly(p2)
p2<-NULL

### Set Negative New Case or Death Counts to 0
cov_states$new_cases[cov_states$new_cases < 0 | is.na(cov_states$new_cases)] = 0
cov_states$new_deaths[cov_states$new_deaths < 0 | is.na(cov_states$new_deaths)] = 0

### Recalculate `cases` and `deaths` as Cumulative Sum of Updated `new_cases` and `new_deaths`
for (i in 1:length(state_list)) {
  cov_subset = subset(cov_states, state == state_list[i])
  
  #### Add Starting Level
  cov_subset$cases = cov_subset$cases[1]
  cov_subset$deaths = cov_subset$deaths[1]}
  
for (j in 2:nrow(cov_subset)) {
  cov_subset$cases[j] = cov_subset$new_cases[j] + cov_subset$cases[j-1]
  cov_subset$deaths[j] = cov_subset$new_deaths[j] + cov_subset$deaths[j-1]
  
  #### Include in Main Dataset
  cov_states$cases[cov_states$state==state_list[i]] = cov_subset$cases
  cov_states$deaths[cov_states$state==state_list[i]] = cov_subset$deaths}

### Smooth New Counts
cov_states <- cov_states |>
  mutate(
    new_cases = zoo::rollmean(new_cases, k = 7, fill = NA, align = 'right') |> round(digits = 0),
    new_deaths = zoo::rollmean(new_deaths, k = 7, fill = NA, align = 'right') |> round(digits = 0))


### Inspect Data Again (Interactively)
p2 <- ggplot(cov_states, aes(x = date, 
                             y = new_deaths, 
                             color = state)) + 
  geom_line() + 
  geom_point(size = .5, alpha = 0.5)

ggplotly(p2)
p2<-NULL

Step 5

### Add Population Normalized (by 100,000) Counts for Each Variable
cov_states$per100k = ifelse(cov_states$cases == 0, 0, round(cov_states$cases / (cov_states$population / 100000), 1))
cov_states$newper100k = ifelse(cov_states$new_cases == 0, 0, round(cov_states$new_cases / (cov_states$population / 100000), 1))
cov_states$deathsper100k = ifelse(cov_states$deaths == 0, 0, round(cov_states$deaths / (cov_states$population / 100000), 1))
cov_states$newdeathsper100k = ifelse(cov_states$new_deaths == 0, 0, round(cov_states$new_deaths / (cov_states$population / 100000), 1))

### Add Naive_CFR Variable = Deaths / Cases
cov_states = cov_states |> mutate(naive_CFR = round((deaths*100/cases),2))

### Create `cv_states_today` Variable
cov_states_today = subset(cov_states, date==max(cov_states$date))

Part 2

Step 6

# Population Density versus Cases
cov_states_today |> 
  plot_ly(x = ~pop_density, y = ~cases, 
          type = 'scatter', mode = 'markers', color = ~state,
          size = ~population, sizes = c(5, 70), marker = list(sizemode='diameter', opacity=0.5))
# Filter out "District of Columbia"
cov_states_today_filter <- cov_states_today |> filter(state!="District of Columbia")

# Population Density versus Cases After Filtering
cov_states_today_filter |> 
  plot_ly(x = ~pop_density, y = ~cases, 
          type = 'scatter', mode = 'markers', color = ~state,
          size = ~population, sizes = c(5, 70), marker = list(sizemode='diameter', opacity=0.5))
# Population Density versus Deaths per 100k
cov_states_today_filter |>
  plot_ly(x = ~pop_density, y = ~deathsper100k,
          type = 'scatter', mode = 'markers', color = ~state,
          size = ~population, sizes = c(5, 70), marker = list(sizemode='diameter', opacity=0.5))
# Adding Hover Info
cov_states_today_filter |> 
  plot_ly(x = ~pop_density, y = ~deathsper100k,
          type = 'scatter', mode = 'markers', color = ~state,
          size = ~population, sizes = c(5, 70), marker = list(sizemode='diameter', opacity=0.5),
          hoverinfo = 'text',
          text = ~paste( paste(state, ":", sep=""), paste(" Cases per 100k: ", per100k, sep="") , 
                         paste(" Deaths per 100k: ", deathsper100k, sep=""), sep = "<br>")) |>
  layout(title = "Population-normalized COVID-19 Deaths (per 100k) vs. Population Density for US States",
                  yaxis = list(title = "Deaths per 100k"), xaxis = list(title = "Population Density"),
         hovermode = "compare")

Step 7

p <- ggplot(cov_states_today_filter, aes(x=pop_density, y=newdeathsper100k, size=population)) + 
  geom_point() + 
  geom_smooth() +
  labs(title="New COVID-19 Deaths (per 100k) vs. Population Density", x="Population Density", y="Deaths per 100k")
ggplotly(p)
`geom_smooth()` using method = 'loess' and formula = 'y ~ x'
Warning: Removed 1 row containing non-finite outside the scale range
(`stat_smooth()`).
Warning: The following aesthetics were dropped during statistical transformation: size.
ℹ This can happen when ggplot fails to infer the correct grouping structure in
  the data.
ℹ Did you forget to specify a `group` aesthetic or to convert a numerical
  variable into a factor?

There does not seem to be a particularly strong correlation between COVID-19 deaths per 100,000 and population density. Looking at this figure, we can see that for increases in population density between ~100 and ~240, there is a decline in deaths, as with population density between ~750 and ~1250. For all other regions, there does not seem to be a clear relationship between the two variables.

Step 8

# Line Chart for naive_CFR for All States Over Time Using `plot_ly()`
plot_ly(cov_states, x = ~date, y = ~naive_CFR, color = ~state, type = "scatter", mode = "lines")
# Linechart for Florida Showing new_cases and new_deaths Together
cov_states |> filter(state=="Florida") |> plot_ly(x = ~date, y = ~new_cases, type = "scatter", mode = "lines", name = "New Cases") |> add_trace(x = ~date, y = ~new_deaths, type = "scatter", mode = "lines", name = "New Deaths") 

The peak in new cases was January 10, 2022 with 84.669k new cases. The peak for new deaths was September 20, 2021 with 445 new deaths. The time between these two dates is approximately 112 days.

Step 9

# Map State, Date, and New Cases to a Matrix
cov_states_mat <- cov_states |> select(state, date, cases) |> dplyr::filter(date>as.Date("2021-06-01"))
cov_states_mat2 <- as.data.frame(pivot_wider(cov_states_mat, names_from = state, values_from = cases))
rownames(cov_states_mat2) <- cov_states_mat2$date
cov_states_mat2$date <- NULL
cov_states_mat2 <- as.matrix(cov_states_mat2)

# Create Heatmap Using plot_ly()
plot_ly(x=colnames(cov_states_mat2), 
        y=rownames(cov_states_mat2),
        z=cov_states_mat2,
        type="heatmap",
        showscale=T)
# Repeat with New Cases Per 100k
cov_states_mat <- cov_states |> select(state, date, newper100k) |> dplyr::filter(date>as.Date("2021-06-01"))
cov_states_mat2 <- as.data.frame(pivot_wider(cov_states_mat, names_from = state, values_from = newper100k))
rownames(cov_states_mat2) <- cov_states_mat2$date
cov_states_mat2$date <- NULL
cov_states_mat2 <- as.matrix(cov_states_mat2)

# Create Heatmap Using plot_ly()
plot_ly(x=colnames(cov_states_mat2), 
        y=rownames(cov_states_mat2),
        z=~cov_states_mat2,
        type="heatmap",
        showscale=T)
# Create a Second Heatmap after Filtering to Only Include Dates Every Other Week
filter_dates <- seq(as.Date("2021-06-15"), as.Date("2021-11-01"), by='2 weeks')

cov_states_mat <- cov_states |> select(state, date, newper100k) |> dplyr::filter((date %in% filter_dates))
cov_states_mat2 <- as.data.frame(pivot_wider(cov_states_mat, names_from = state, values_from = newper100k))
rownames(cov_states_mat2) <- cov_states_mat2$date
cov_states_mat2$date <- NULL
cov_states_mat2 <- as.matrix(cov_states_mat2)

# Create a Heatmap Using plot_ly()
plot_ly(x=colnames(cov_states_mat2), y=rownames(cov_states_mat2),
             z=cov_states_mat2,
             type="heatmap",
             showscale=T)

Step 10

# Specified Date

pick.date = "2021-10-15"

# Extract Data for Each State by its Abbreviation
cov_per100 <- cov_states |> filter(date==pick.date) |> select(state, abb, newper100k, cases, deaths)
cov_per100$state_name <- cov_per100$state
cov_per100$state <- cov_per100$abb
cov_per100$abb <- NULL

# Create Hover Text
cov_per100$hover <- with(cov_per100, paste(state_name, '<br>', "Cases per 100k: ", newper100k, '<br>', "Cases: ", cases, '<br>', "Deaths: ", deaths))

# Set Up Mapping Details
set_map_details <- list(
  scope = 'usa',
  projection = list(type = 'albers usa'),
  showlakes = TRUE,
  lakecolor = toRGB('white'))

# Make Sure Both Maps Are On the Same Color Scale
shadeLimit <- 125

# Create Map
fig <- plot_geo(cov_per100, locationmode = 'USA-states') |> 
  add_trace(
    z = ~newper100k, text = ~hover, locations = ~state,
    color = ~newper100k, colors = 'Blues')
fig <- fig |> colorbar(title = paste0("Cases per 100k: ", pick.date), limits = c(0,shadeLimit))
fig <- fig |> layout(
    title = paste('Cases per 100k by State as of ', pick.date, '<br>(Hover for value)'),
    geo = set_map_details)
fig_pick.date <- fig

# Map for Today's Date

# Extract Data for Each State by its Abbreviation
cov_per100 <- cov_states_today |>  select(state, abb, newper100k, cases, deaths)
cov_per100$state_name <- cov_per100$state
cov_per100$state <- cov_per100$abb
cov_per100$abb <- NULL

# Create Hover Text
cov_per100$hover <- with(cov_per100, paste(state_name, '<br>', "Cases per 100k: ", newper100k, '<br>', "Cases: ", cases, '<br>', "Deaths: ", deaths))

# Set Up Mapping Details
set_map_details <- list(
  scope = 'usa',
  projection = list(type = 'albers usa'),
  showlakes = TRUE,
  lakecolor = toRGB('white'))

# Create Map
fig <- plot_geo(cov_per100, locationmode = 'USA-states') |> 
  add_trace(
    z = ~newper100k, text = ~hover, locations = ~state,
    color = ~newper100k, colors = 'Blues')
fig <- fig |> colorbar(title = paste0("Cases per 100k: ", Sys.Date()), limits = c(0,shadeLimit))
fig <- fig |> layout(
    title = paste('Cases per 100k by State as of', Sys.Date(), '<br>(Hover for value)'),
    geo = set_map_details)
fig_Today <- fig

### Plot together 
fig_combined <- subplot(fig_pick.date, fig_Today, nrows = 2, margin = .05)

fig_combined

There are obvious differences in CFR for the two figures. First, we can immediately see a difference in the amount of cases for the two dates based on the colors of each figure, with the 2024 map being almost completely a light shade of blue. The 2021 map, on the other hand, is much more heterogenous, with states like Alaska, Montana, and West Virginia having the most cases.